Professional Documents
Culture Documents
Analytic Surfaces
Analytic Surfaces
Analytical Surface
Deformation Theory
for Detection of the Earth's Crust Movements
With 20 Figures
" Springer
Dr. Yiiksel Altmer
Bundesamt fUr Kartographie und Geodasie
Richard-Strauss-Allee 11
D-60598 Frankfurt/Main
Germany
E-mail: altiner@ifag.de
ISBN 978-3-642-08510-9
This work is subject to copyright. All rights are reserved. whether the whole or part of the material is
concerned. specifically the rights of translation. reprinting, re-use of illustrations. recitation,
broadcasting. reproduction on microfilm or in other ways, and storage in data banks. Duplication
of this publication or parts thereof is permitted only under the provisions of the German Copyright
Law of September 9.1965. in its current version, and permission for use must always be obtained from
Springer-Verlag. Violations are liable for prosecution act under German Copyright Law.
© Springer-Verlag Berlin Heidelberg 1999
Originally published by Springer-Verlag Berlin Heidelberg New York in 1999
Softcover reprint of the hardcover 1st edition 1999
The use of general descriptive names, registered names. trademarks, etc. in this publication does not
imply. even in the absence of a specific statement. that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
Due to plate motions, tidal effects of the Moon and the Sun, atmosphe-
ric, hydrological, ocean loading and local geological processes, and due
to the rotation of the Earth, all points on the Earth's crust are sub-
ject to deformation. Global plate motion models, based on the ocean
floor spreading rates, transform fault azimuths, and earthquake slip
vectors, describe average plate motions for a time period of the past
few million years. Therefore, the investigation of present-day tectonic
activities by global plate motion models in a small area with complex
movements cannot supply satisfactory results.
The contribution of space techniques [Very Long Baseline Interferome-
try (VLBI); Satellite Laser Ranging (SLR); Global Positioning System
(GPS)] applied to the present-day deformations ofthe Earth's surface
and plate tectonics has increased during the last 20 to 25 years. Today
one is able to determine by these methods the relative motions in the
em to sub-em-range between points far away from each other.
This high accuracy achievable using space techniques requires rethin-
king of the theoretical foundations of deformation analyses to be app-
lied. These analyses constitute the subject of chapter 2 of this book,
in which an analytical theory of internal and external surface defor-
mations is considered. Their application requires geometric surface
modelling, a survey of which is given in chapter 3. Finally, Chapter
4 has been dedicated to application of the model to the velocities de-
rived within the CRODYN'94 and CRODYN'96 GPS campaigns in
the Adriatic Sea area. In this case it is only a methodical application
the results of which cannot be considered as sufficiently certain. More
satisfactory statements can be expected only after several additional
GPS campaigns in this area.
VIn Preface
1 Introduction 1
2 Deformation Theories 7
2.1 Coordinate Systems . . . . . . . . . . . . . . . 7
2.1.1 Cartesian and Curvilinear Coordinates. 7
2.1.2 Surface Normal Coordinates. 9
2.2 Three-Dimensional Deformations 10
2.2.1 Preliminary Remarks. . . . . 10
2.2.2 General Fundamentals . . . . 11
2.2.2.1 Isoparametric Representation. 11
2.2.2.2 Lagrangian Representation 11
2.2.3 Special Deformation Measures 14
2.3 Surface Deformations. . . . . . . . . . 15
2.3.1 Preliminary Remarks. . . . . . 15
2.3.2 Internal Surface Deformations. 15
2.3.2.1 Isoparametric Representation. 15
2.3.2.2 Computations on the Basis of the Gaus-
sian Surface Representation. . . . . . .16
2.3.2.3 Computations in External Surface Nor-
mal Coordinates . . . . . . . . . . .. . 19
2.3.2.4 Computations in Ellipsoidal Coordinates 20
2.3.3 External Surface Deformations . . . . . . . . . 22.
2.3.3.1 Computations on the Basis ofthe Gaus-
sian Surface Representation . . . . ... 22
2.3.3.2 Computations in External Surface Nor-
mal Coordinates . . . . . . . . . . . . 29 .
2.3.3.3 Computations in Ellipsoidal Coordinates 32
2.3.4 Deformations and Stresses . . . . . . . . . . . .. 33
x Contents
3 Geometric Modelling 39
3.1 General Fundamentals 39
3.2 Interpolation by Polynomials 42
3.3 Interpolation by Collocation . 44
3.4 Subdivision of a Surface by Finite Triangle Elements 49
3.4.1 Preliminary Remarks . . . . . . 49
3.4.2 Triangle Elements . . . . . . . . . . . . . . . 50
3.4.3 Discrete Deformation Measures . . . . . . . . 52
3.4.4 Determination of the Change of the Bend Bet-
ween the Neighbouring Triangles . . . . . . . . 54
4 Application 57
4.1 Deformation of the Earth's Crust 57
4.2 Geodetic Contributions to Deformation Analyses 62
4.3 Tectonic Development of the Adriatic Sea Area 63
4.4 Geometric Modelling . . . . . . . . . . . . . . . . 64
4.4.1 Interpolation Methods Applied . . . . . . 64
4.4.2 Explanations of the Graphic Representations 66
4.5 Surface Deformations in the Adriatic Sea Area 70
4.5.1 GPS Observations . . . . . . . . . . . 70
4.5.2 Processing Strategy of the GPS Data 73
4.5.3 Data Processing . . . . . . 73
4.5.4 Checking of Data Quality . . . . . . . 74
4.5.5 Multi Campaign Solutions . . . . . . . 76
4.5.5.1 Absolute Coordinate Solution. 78
4.5.5.2 Relative Coordinate Solution 78
4.5.6 Internal Surface Deformations .. 79
4.5.7 External Surface Deformations . 80
4.5.8 Figures of the Adriatic Sea Area 81
4.6 Summary . . . . . . . . . . . . . . 88
Bibliography 89
Index ............................ 98
Chapter 1
Introduction
aE{I,2,3} (1-3)
the qa, ir
being generally ellipsoidal coordinates. From (1- 3) the
'displacement coordinates'
(1-4)
can be formed. These represent the displacement field of the region
F of the Earth's surface to be investigated (1-1), which is generally
better the denser the field of the observation stations P is selected. It
must be observed here that the station density to be selected, and thus
the quality of representation of the displacement field, is not only a
function of the height variations within F. Also in relatively flat areas
an appropriately dense distribution of observation stations must be
given to detect possible short wavelength displacement components.
Owing to (1-2, 3) Pm - Pn
linear elongations
(1-6)
Deformation Theories
Xi . in S , (2.1-1)
are used. In three-dimensional Euclidean as well as in two-dimensional
Riemannian spaces are introduced
general curvilinear coordinates:
three-dimensional:
qa, aE{1,2,3}, metric tensor gab ; (2.1-2a)
two-dimensional sur/ace coordinates:
UCX, a E {1, 2} , metric tensor / cx{3 • (2.1-2b)
The coordinate geometry on the basis of these systems is represented
according to Heitz (1988).
For the coordinate geometry the
general differential equations of the coordinate transformati-
ons qa {:} ir :
q- C,ab = rdab it,d _ fcde ijd,a ije,b' ijc:=
,d 8ijc/8 qd (2.1-3a)
8 Chapter 2. Deformation Theories
Xkd
,
= Xk ' d := 8Xk . /8 q d =: Ck . d , qd.,t := 8qd/8xi . =: C~l. ., (2.1-3c)
(2.1-4a)
a E {1,2} :
(2.1-4b)
(2.1-5a)
a E {I, 2}
(2.1-5b)
2.1. Coordinate Systems 9
(2.1- 6a)
(2.1-6d)
(2.1- 8a)
a S (1)
p - System (A" <1>, h)
qa_ System (A" <1>, H)
_ - 4 r " - - - - - - S (t)
EIIi'W~
t and f (2.2-1b)
in the
isoparametric representation
with the
deformation tensor:
(2.2-4b)
For the
metric tensors
the relation
( Xi a Zi b
"
+ Xi b Zi a
"
+ Zi a Zi b)/2
"
(2.2-5c)
(2.2-6a)
(2.2-7a)
2.2. Three-Dimensional Deformations 13
the
Cartesian deformation tensor
D··
tJ. := ctJ
". + Vik' Vik'J. /2
t. = D Jt.
·· (2.2-7b)
with
(2.2-7c)
(2.2-7d)
v... -
tJ . -
c ..
c"tJ. - <,tJ. (2.2-7e)
2DtJ.
· · dx t'. dx J.' (2.2-7f)
c ..
<"J. = C"k
tJ . dk . (2.2-8a)
and with
Vi. := dZi./ dt ,
W·t . (2.2-8b)
(2.2-11a)
(2.2-11b)
(2.2-11c)
where
+1 if i, j, k are cyclic,
- 1 if i, j, k are anticyclic, (2.2-11d)
o if i, j, k are acyclic,
are the permutation symbols for the Cartesian coordinates. For the
permutation symbols of the ellipsoidal coordinates see (2.3-27d).
With (2.2-11a-c) one obtains for the
volume dilatation (-depression):
(2.2-12a)
2.3. Surface Deformations 15
(2.3-3a)
with the
surface deformation tensor:
(2.3-3b)
For the
metric tensors
the relation
(2.3-5c)
(2.3-6a)
the linear surface deformation tensor
applies sufficiently precisely
n E {I, 2} , (2.3-7c)
which are identical with the directions of the principal axes of the
linear deformation tensor cO/{3, one obtains starting from (2.3-7a) as
the conditional equation the quadratic equation
ao v 2 + al v + a2 = 0, ao = El2 122 - C22 h2 ,
(2.3-7d)
al = Ell 122 - C22 111, a2 = Cll 112 - Cl2 111
The two solutions of the quadratic equation generally read
(2.3-7f)
VI -l- 0 ,
(2.3-7h)
and
(2.3-7i)
(2.3-7k)
result.
2.3. Surface Deformations 19
[u l , u2 , H( UC¥) ja ,
[ ul ( UC¥), u2( UC¥), fIe UC¥) ja . (2.3-8a)
(2.3-8b)
(2.3-9a)
(2.3-9b)
and it is
(2.3-10a)
COt(3 = [q~Ot (z~ + r~f q~(3 zf) + q~(3 (z~Ot + r~f q~Ot zf) 19cd/ 2 (2.3-10c)
results. The first derivations of the qC according to the surface coordi-
nates result on the basis of (2.3-8a) to
(2.3-10d)
For the internal surface deformation measures here also apply
(2.3-7a-k).
aE{1,2}, (2.3-11a)
(2.3-11c)
gab = (Rl + H)2 cos 2 </> 6! 6l + (R2 + H)2 c5~ c5~ + c5~ c5~ (2.3-12a)
with
The first partial derivatives of the metric tensor (2.3- 12a) resulting
from taking >., </> and H read
911 ,1 0
911,2 -(R1 + H)(R2 + H) sin 2</>
911,3 2(R1 + H) cos2 </>
922 ,1 o (2.3-12b)
922,2 3R 2(R 2 + H)( e" /V)2 sin 2</>
922,3 2(R2 + H)
933,1 933,2 = 933,3 = 0
with
e" 2 = 2nd numerical eccentricity of the ellipsoid of revolution (2.3-
12a) .
From (2.3-12) one obtains the Christoffel symbols of the 2nd kind
needed in (2.3-9c)
r~b = 9cd (9ad,b + 9bd,a - 9ab,d) / 2, (2.3-13a)
as follows:
rb = -(R2 + H) tan </> / (R1 + H) ,
rb = 1/ (R1 + H) ,
r 121 --r 12'
1
r 31
1 - r1
- 13'
n1 = (R1 + H)
sin2</> /2(R2 + H) ,
n2 = 3R2(e" /V)2 sin 2¢ / 2(R2 + H) , (2.3-13b)
one obtains the metric tensor fOt{3 in the first point of time t in the
isoparametric surface representation (2.3-2)
fOt{3 = [(Rl+H)2cos24>+(H,1)216~6b (2.3-16)
+(6~ 6~+6~ 6b)H,1 H,2+[(R2+H)2+(H,2)2l6~ 6~.
Thus, all relevant quantities have been determined to be able to com-
pute the metric tensor JOt{3, (2.3-10a) for the second point of time t,
and the linear surface deformation tensor fa{3, (2.3-10c). The linear
surface elongation and the surface dilatation (-depression) result
according to (2.3-7).
(2.3-18b)
Cijk. [Xj,1 Xk,2 Zi,OI{3
DL L-L (2.3-18c)
the
24 Chapter 2. Deformation Theories
(2.3-19b)
and the
Gaussian curvature
(2.3-19c)
defined here as geometrical mean, by
n E {I, 2} , (2.3-20a)
which are identical with the directions of the principal axes of the
second fundamental tensor L a {3, one obtains starting from (2.3-19a)
as conditional equation the quadratic equation
V(1) -+ 0 ,
(2.3-20d)
Ib2 1 ~ Ibll and Ibol ~ Ibll
2.3. Surface Deformations 25
and
(2.3-20g)
n E {I, 2} (2.3-20h)
These results (2.3-20a-i) are formally true also for the deformed sur-
face P with
= f a {3 8La{3 + La {3 6r{3
and the
change of the Gaussian curvature
(2.3-22b)
(2.3-23a)
2.3. Surface Deformations 27
(2.3-24a)
(2.3-24b)
(2.3-25a)
(2.3-25b)
(2.3- 26a)
Analogically, the
tensor of rotation (2.3-25a, b) with (2.3-22b) can be described as
follows:
(2.3-26b)
The
infinitesimal rotation vector pertinent to the tensor of rotation
28 Chapter 2. Deformation Theories
Xi,~ == Xi,a q~ ,
and
(2.3-29b)
(2.3-31a)
(2 .3-31c)
(2.3-31d)
zi,b = Ci.d ( z,bd + rdef q,be Zf )=.. Ci.d o{3b ( z,iJd + rdef q,jje Zf ) . (2.3-32b)
especially:
ql'_
,a
= 01'ex ,
q3,a -- H ,0',
(2 .3-34c)
qd..,3 cd1. n'1. = nd = gcd cabc qa.,1 qb_,2 f- 1 / 2
with
f
J Ot(J = q,aa q,iJb gab , (2 .3-34d)
and
ni. = bi,3 = normal of the reference surface of the pa - system.
For the transformation matrices between the pa-system and the qa-system
inverse to this applies
(2.3- 35a)
especially
(2 .3- 36b)
32 Chapter 2. Deformation Theories
(jc_
;(3
= (1/2)gabc [qg_,b (zd_,a + rdej qe_,a zj)gdg ] ;/3 .
The result must be identical with (2.3-28b) if there the Cartesian
coordinates are eliminated. In this way first
(2.3-37b)
c ( z d-{3-
qfi
',Q
+ rdejqijz(3-
e
"
j + rdejq e-(3-z j + rdejoqijq{3-z
,0 ,
e
,
0
,
j) gcd
(2.3-37c)
Xi,3 Ci3.
a f- 1/ 2 Ca (2.3-39b)
P,1 11 1,
P,2
a
(f12 , -f11 , ot (f f11)-1/2 ,
a C3 ,
P,3
with
(2.3-39c)
34 Chapter 2. Deformation Theories
With this the three Cartesian components ell., e12., e22. can be assu-
med as given. A further precondition consists of the disappearance of
the surface stresses
(2.3-41b)
A direct result is
e12. =0 (2.3-41e)
2.3. Surface Deformations 35
(2.3-41g)
(2.3-42a)
(2.3-42b)
with
(2.3-43b)
(Ti3,3 = -2/1 {ci1,1 + Ci2,2 + [v /(1 - v)] ( q,l Dil. + q,2 Di2. )} . (2.3-43c)
36 Chapter 2. Deformation Theories
So far all formulas are rigorous. For the further evaluation of O"i3,3
the approximations are done in the neighbourhood of the test
point P
F == tangential plane of F in P ,
(2.3-44a)
F = curved surface,
which are justified with the assumption of small displacements (2.3-
6a). Hence, the following simplifications apply
f = I ,
Xi,c.{3 =0 , (2.3-44b)
Lc./3 =0 j
(2.3-44c)
Zi,3 =0 , Zi,3c. =0 j
fc./3 = oc./3. + zc.,/3 + z/3,c. = oc./3. + 2 Ec./3. ,
If the axes 1., 2. are identified with the principal directions of curva-
ture, then it applies that
(2.3-45c)
6H = iI - H , (2.3-47b)
Geometric Modelling
where
apq coefficients
functions to be determined of u~, {Yq, u~} , (3.1-5b)
Y; trend function
known function of u~ . (3.1-5c)
[u 1 , U2 , H(ucx)]a ,
[ ii,l( U CX ), ii,2( U CX ), H( ucx)]a , (3.2-2a)
(3.2-2b)
For the region F of the Earth's surface and the displacement field
observed, one starts from analytical functions in the form of a
Polynomial for the height q3 == H :
n~ n~
H(u CX ) ~ L L hn1n2 (U1tl (U 2 t2 (3.2-3a)
nl=On2=O
and of
Polynomials for the displacement coordinates za :
n~ n;
za( UCX) ~ L L c~ln2 (u 1 tl (u 2 )n2 • (3.2-3b)
nl=On2=O
(3.2-3c)
is done on the basis of the H - and za-values (3.2-2a, b), i.e. gener-
ally by a formal application of the adjustment according to the least
squares method. After the determination of the coefficients (3.2-3c)
all measuring quantities treated in sections 2.3.2 and 2.3.3 can be
computed with regard to the internal and external surface deformati-
ons.
The degree of the polynomials
n' = n~ + n~ (3.2-4a)
If n~r is the larger and nkl the smaller one of the degree numbers ni, n2
the number of coefficients amounts to:
- - -T (3.3-1a)
YP := YP - Yp ,
in the form
(3.3-3c)
nq nq nq
(3.3-4a)
3.3. Interpolation by Collocation 47
(3.3- 4b)
for the coefficients apq , which when applied to (3.3-3c) read fully:
nq
Qp, Qq are the foot points of Pp, Pq in the reference surface of the
surface normal coordinates. A representation of the covariance functi-
on that is in general well adapted to the specific requirements can be
achieved with exponential functions:
(cf. Heitz 1968). The first exponential function describes the cova-
riance for the near zone and the second one must ensure a sufficiently
strong decrease in the distant zone in order to avoid singularity of the
48 Chapter 3. Geometric Modelling
Yp,a ... =L apq,a ... Yq apq,a ... = L C;/ Cpl,a .... (3.3-7b)
q=l q=l
For the covariance function (3.3-5b), the partial derivations result to:
(3.3-7c)
[cld~ exp( -d1 S;q) + c2d~ exp( -d2 S;q)] S;l,a S;I,{3
-[ c1d1 exp( -d1 S;q) + C2d2 exp( -d2 S;q)] S;I,a{3 ;
2
Spl,a{3 .=
• £l2 2 / £l a £l {3
Spl uU uu
U .
(3.3-8a)
3.4. Subdivision of a Surface by Finite Triangle Elements 49
and therefore
and
(3.4-2a)
(3.4-2b)
The coefficients C , c{3 and ca , c~, respectively, are computed via the
conditional equations in the corners Pq
(3.4-2c)
(3.4-3)
(3.4-4c)
if q + n > 3, then q + n =: q + n - 3 ,
52 Chapter 3. Geometric Modelling
3 3
C = L[bq/D] Hq , Cf3 = L[(bf3)q/ D] Hq ,
q=l q=l (3.4-5a)
3 3
ca = L[bq/ D] (za)q , Cp = L[(bf3)q/ D] (za)q .
q=l q=l
The partial derivations of the heights and displacement coordinates
according to the test point coordinates uf3 result from (3.4-2b) with
the mode of designation (3.3-7a) in:
n E {1, 2} (3.4-9b)
B~__nt-i._ _-:::;~ C
Fig. 3.2. The change of the bend between the triangles describes the
changes of the external form of the investigation area
Chapter 4
Application
Some models dealing with global plate motions were published in (De-
Mets et al. 1990, 1994; Argus and Gordon 1991; Seno et al. 1987,
1983; Wilson 1993a, 1993b). These models describe plate motions by
a rotation vector (Euler Vector) of the plate based on a pole of rota-
tion. The NUVEL-1 model describes motions between 14 major rigid
plates relative to the fixed Pacific plate (Demets et al. 1990). The
NNR-NUVEL1 model (no net rotation) gives absolute angular velo-
cities of the plates (Argus and Gordon 1991). Due to recent revision
in the paleomagnetic time scale, the angular velocities of these earlier
models are multiplied by a recalibration factor of 0.9562 (DeMets et al.
1994). These rescaled models are known as NUVEL-1A and NNR-
NUVEL1A global models. As an example, the angular velocities of
the lithospheric plates in the NUVEL-1A model are given in Table 4.1
(Demets et al. 1994).
:50
-50
Zi.s = Rwp [
COS ¢p sin Ap sin ¢s - sin ¢p cos ¢s sin As
cos¢s cos¢ssin As -.cos¢psin¢scosAp
1, (4.1-1)
cos ¢p cos ¢s sm(As - Ap)
where (A, ¢)s are the coordinates of the station on the plate, (A, ¢)p
are the coordinates of the rotation pole of the plate, and R is the
spherical Earth's radius. w describes the angular velocity of the plate.
With
Xi.s(t) = Xi.s(tO) + Zi.s(t - to) (4.1-2)
one obtains for the observation epoch t the corrections of the stati-
on's coordinates due to the global plate motion relative to a reference
epoch to. Gravitational attraction of the Moon and the Sun on
points on the Earth's crust, illustrated in (Fig. 4.3), is not constant
because the Moon and the Sun change position due to the rotation
of the Earth. This changing attraction of the Moon and the Sun is
the main cause of oceanic and solid Earth tides. For the modelling
of the tidal effects (see, e.g., Melchior 1983; Heitz 1980-1988; Va-
nicek and Krakiwski (1986); Sovers and Jacobs 1994; McCarthy 1996).
4.1. Deformation of the Earth's Crust 61
~I~'~'
-
!!. ..
- II:·
moon Greenwich
meridian
'--_ _ _ rotation angle
,, UT1
90' western equator ,.
longitude cP vernal equinox
5mm/ ~
Fig. 4.4. The Adriatic Sea area is a promontory of the African plate
and movement of this promontory in a north-northwest direction is
about 5 mm/yr. AR: Arabia; AN: Anatolia; BS: Black Sea; AE: Aegea;
AD: Adriatic block; PA: Pannonian basin; TY: Tyrrhenian basin; BA:
Balearic basin; BB: Bay of Biscay; GM: Gibraltor-Morocco block;
RGs: Rhine Graben system (Mantovani et al. 1992)
Hungary
45
'1, '
. 0;;" . 44
0",.
1'0
. &
.;. ~o
14 IS 16 17 18 19
Fig. 4.5. Earthquakes occurred in the last 500 years III the Adriatic
Sea area (Skoko and Mokrovic 1998)
66 Chapter 4. Application
as described in sections 3.1 and 3.3, was used. The heights and velo-
cities were also interpolated by the spline method for the comparison
of both results. In this case, the differences between the coordinates
interpolated by the spline and collocation methods were very small
and can be disregarded.
The method of the subdivision by finite triangle elements or trian-
gulation is well suited for the computation of internal deformation
measures [cf. (3.4-7a)], as well as for the densification of observation
stations, because in this case only the functions themselves are conti-
nuous at the triangle sides, but not their derivatives. However, for the
determination of the external deformation measures, the collocation
method can be applied in a considerably simpler and clearer way, sin-
ce it provides interpolation functions that can be differentiated in the
whole area of interpolation two times continuously according to the
surface coordinates uOt. Their application then allows us to compute
directly all internal and external deformation measures on the basis of
the formulas presented in sections 2.3.2 and 2..4-3.
Computation of the internal deformation measures was performed on
the basis ofthe formulas (2.3-4a-10d) from sections 2.3.2.2 and 2.3.2.3
taking into account the ellipsoidal coordinates [cf.section 2.3.2.4]. In
contrast to this, the formulas (2.3-17a-37c) from sections 2.3.3.1,
2.3.3.2 and 2.3.3.3 were used to compute the external deformation
measures.
Z = Z{"\, ¢) . ( 4.4-1 a)
4.4. Geometric Modelling 67
8H
l
R 8qS
p----R1 COSqS8A~
Fig. 4.6. Representation of displacement coordinates
( 4.4-3a)
68 Chapter 4. Application
oH('x, 4» ( 4.4-3d)
(4.4-4a)
which can be represented according to Fig. 4.7. With this q(n) is, as a
rule, to be marked off in an enlarged form starting from the unit circle
additionally represented
instead of ( 4.4-4a) the line length in the drawing. Further, the tangent
vector to the contour in the test point P('x, 4>, H)
( 4.4-4d)
( 4.4-5a)
(4.4-5b)
( 4.4-5c)
and
(4.4-5d)
( 4.4-5e)
the relations
( 4.4-5f)
70 Chapter 4. Application
( 4.4-5g)
the vectors
(4.4-6b)
are appropriately represented. For this purpose only the original di-
rection vectors r(n) are used; a representation of the directions of mean
curvatures f(n) does not seem appropriate. In contrast to the positive-
ly defined distance conditions the changes of curvature may be positive
and negative.
Hungary
Siovenio
Table 4.3. The individual reference frames of the CODE orbits used
in the processing
The coordinates of the IGS sites (International GPS Service for Geo-
dynamics) Graz (Austria), Matera (Italy), Zimmerwald (Switzerland)
and Wettzell (Germany), which are included in the processing to de-
fine the datum of the reference frames, were shifted to the individual
observation epochs of the CRODYN'94 and CRODYN'96 campaigns
using the coordinates and velocity values published in the IERS Tech-
nical Notes for the ITRF92 and ITRF94.
9
D Lat.
- Lon .
DH
~ 5
E
.§.
4
n +---'---
158 159 160
Session
a laI.
6 • Lon.
011
4
e
-5.
-----
2
0+--=-=
253 254 255
Session
The normal equations from each daily solution were combined using
the ADDNEQ program of the Bernese software to create a free cam-
paign solution with 1 m a priori standard deviations for all station
coordinates (Table 4.6). On the basis of a 7 Helmert transformation
parameters , the station coordinates of the daily solutions of each in-
dividual campaign were compared separately with the results of the
free campaign solution. The standard deviations for the station coor-
dinates derived by the 7 Helmert transformation parameters are less
than ±3 mm for the horizontal components and approx. ±5 mm for
the vertical component (Fig. 4.9 and 4.10).
the campaigns have the same quality and are uncorrelated, the obser-
vations of the coordinate components y can be modelled as a linear
function of time y + e = at + b = XB with D(y) = 0'51 (Brock-
mann 1996). The unknown parameters B = (a, b)T can be computed
according to the Gauft-Markoff Model (XTpxfj = XTpy) with its
variance-covariance matrix D(fj) = &2(XT p X)-I, with n number of
observations, u number of unknown parameters, f3 vector of unknows
(ux1), X matrix of coefficientes with full rank rgX (nxu), y vector
of observations (nxl), e vector of residuans (nxl), P positive definite
weight matrix, 0'2 variance of unit weight. In this case, the modelling
of y as a linear function of time can be interpretated as a model of
linear regression.
For the computation of the accuracy for the station coordinates the
transformation of the Cartesian error components to the ellipsoidal
components were made by means of the position vector (xiJ
X =[ -
-sin A I N 1
cos A sin </> I N 2 -
cos A I N 1
sin A sin </> I N 2
0
cos </> I N 2
1 (4.5-1b)
cos A cos</> sin A cos </> sin </>
L
48
46
.,
.c....
"tl
d
'-l
44
Longitude
50
1WEnm1.
48
.~
,.....
j~
46
~
~
::::d
'-l
44
:!
i"j
42
1
I*
u
Ii
40 ~
10 15 20
Longitude 2.0 em / yeor
• Croz
47
\
46
] 45
...
d
"'I
44
!
I~
43
I
~
42 L -__~____L-__~____L-__~____L -_ _~_ _~~
12 14 16 18 20
Longitude
• CrOl
47
46
I~oo.
44 ~
'!
i!
I
1
43 ~
~
~
42 ~ __ ~ ____- L____ ~ ____ ~ ____ ~ ____L-__ ~ ____ ~
:§;
12 14 16 18 20
Longitude
46
45
44
..,
N
~
110
en
en
"i
Q.
Jl
I
43
~
.
,.,.
<:
Ii
CI
:..:
m
.....
~
14 15 16 17
Longitude
46
.................. ,
, , ............. '\ \
, .
·0 ,. •Brosnik'
............... .
........
" ...
" ...
,\"'"
'\
'\
,
, \ ......:
'\ ,""\
"""'l,
......... , \ 1 l'
......Nll...........,-...,,"""'\""",. · ~~" ~ ... "'\ l,\'l""\
44
-
IX)
01
01
I
a..
JlI
co
43 .,"0c
l-
e>
14 15 17
Longitude .25 ~train/year
Fig. 4.16. Surface elongations. The arrows indicate the direction and
magnitude of extension. The lines show the direction and magnitude
of compression. The results are given in J.lstrain
4.5. Surface Deformations in the Adriatic Sea Area 87
Blegos
46
45
o Brusnik
C»
"t!
~
;.>
.~
;.> I
t3
"4 ''""
0
44 I
~,\cono
Sea
43
:.. a:I
O· ·0 ' :51 ......
'<=:r=:oti\~~ 1 c:E - ·\..? , ~~ ~
14 15 16 17
Longitude
4.6 Summary
The frequencies of earthquakes occurences in the Adriatic Sea area
show that the seismic activity is concentrated especially in the nor-
thwestern part and southern part of the investigation area. The last
earthquakes in Croatia and Slovenia in 1996 and in 1998 can be seen
as supporting this statement. A comparison of the results derived on
the basis of the analytical surface deformation theory with the results
of the seismological and geological investigations in the same area in-
dicates a good coincidence between two results. The concentration
of the earthquakes in the Adriatic Sea area covers the area with the
maximum surface elongations computed within the CRODYN'94 and
CRODYN'96 GPS campaigns.
The maximum deformations derived in the north- western part of the
investigation area with an extension of 2 mm/yr per 10 km in the
north-western and south-eastern directions could be seen as an indi-
cation of the magnitude 5.8 earthquake in Bovee (Slovenia) on April
12, 1998. A magnitude 6.0 earthquake in Ston occurred on September
5, 1996 in the southern Croatia, also an area where maximal extensi-
ons was observed.
Preliminary results obtained from two GPS campaigns support the
opinion that the Adriatic Sea is a promontory of the African plate
(Fig. 4.15 and 4.16). A south-north compression of 1 to 2 mm/yr
per 10 km across the regions of extensions is consistent in direction
and magnitude with the movement of the African plate in this area
(Mantovani et al. 1992).
Concidering the short duration of the investigation period, the que-
stion of whether the Adriatic Sea is a promontory of the African
plate or exists as an independent microplate cannot be answered se-
riously. Moreover, for the elimination of the model errors and influ-
ences of local geological processes on the station's coordinates and
velocities, the results of a third GPS campaign are to be taken in-
to consideration. Furthermore, for a seriously assertion about the
present tectonic activities in the Adriatic Sea area by means of the
GPS, a long investigation period from 8 to 10 years is to be covered.
Bibliography
ETRF, 57 lithosphere, 58
Euclidian space, 1 local stress-deformation, 33
Euler vector, 59 local system, 33
exponential function, 47 Cartesian, 34
extension, 79
mean curvature, 24
Finite Element Method, 49 measure
first fundamental tensor, 22, 23 primary, 4
function method
covariance, 47 approximation, 41
exponential, 47 collocation, 66
spline, 42 interpolation, 41
trend, 41 spline, 66
metric
GauB-Markoff Model, 77 geographical ellipsoidal, 20
Gaussian curvature, 24 metric tensor, 7,9, 12, 16, 19,33
geometric modelling, 39 first derivation, 21
geometry transformation, 22
Euclidean, 7 microplate, 64
external, 1 mid-oceanic ridge, 58
internal, 1
GPS, 2 NNR-NUVEL1,59
GPS network, 63 NNR-NUVEL1A, 59
gravitational attraction, 60 normal curvature, 23
normal unit vector, 54
Hooke's law, 34
NUVEL-1, 59
IGS, 73 NUVEL-1A, 59
stations, 58
interpolation ocean floor spreading, 58
collocation, 42, 44
parameter
polynomial, 42
elacticity, 34
spline, 42
partial derivation
interpolation function, 40
first, 29
analytic, 41
second, 29
isotropic, 34
permutation symbol
ITRF, 57, 73
Cartesian, 14
kinematic model, 62 ellipsoidal, 28
plate
law of transformation, 22 African, 63
Least-Squares-Postulate, 46 Eurasian, 63
linear plate boundaries, 60
deformation plate motion, 57
tensor, 12, 20 plate tectonics, 3
interpolation, 41 polynomial degree, 43
100 Index
tangential plane, 36
test point, 33
Theorema egregium, 24
tidal effect, 57, 61
Springer
and the
environment
At Springer we firmly believe that an
international science publisher has a
special obligation to the environment,
and our corporate policies consistently
reflect this conviction.
We also expect our business partners -
paper mills, printers, packaging
manufacturers, etc. - to commit
themselves to using materials and
production processes that do not harm
the environment. The paper in this
book is made from low- or no-chlorine
pulp and is acid free, in conformance
with international standards for paper
permanency.
Springer